We propose to continue the follow-up of over 116,000 women who in 1989 were enrolled in a prospective study to evaluate potentially modifiable risk factors for major health problems among younger women. In this unique cohort, the Nurses' Health Study II (NHS II), exposure information has been collected at two-year intervals in early adult life, and we have maintained over 90% response to follow-up questionnaires. The proposed specific aims build upon and extend the original objectives. Specifically, we will tet the following hypothesis for breast cancer and melanoma, the two most important malignancies in this age group, and will quantify and characterize associations that are observed: 1) higher intakes of specific carotenoids, dietary fiber, and unsaturated fat during adolescent and early adult life reduce risk of premenopausal breast cancer, whereas higher dietary glycemic load and intake of saturated fat increase risk; 2) higher levels of physical activity reduce risk of breast cancer (for this aim we will evaluate contributions of different activity levels during different stages of life, and assess the degree to which error in measuring activity or recall bias can account for major inconsistencies in the literature); 3) current use of lower- dose oral contraceptives does not increase breast cancer risk, but long-term use before first pregnancy increases risk; 4) breast-feeding reduces risk of breast cancer among premenopausal women; 5) long duration of oral contraceptive use and low intake of specific antioxidants increase the risk of melanoma; and 6) higher plasma levels of IGF-1 increase breast cancer risk. In the proposed follow-up, we will continue to mail questionnaires to participants at two year intervals to update information on the primary exposures and relevant covariates, and to ascertain incident cases of cancer and other major illness. For all reported cancers, we seek medical records for confirmation and more detailed characteristics. Through 2003, we expect to ascertain 1,685 cases of invasive breast cancers and 301 invasive melanomas, thus providing substantial statistical power to address the above hypotheses. A nested case-control study of IGF-1 in relation to breast cancer will utilize the blood specimens currently being collected from approximately 30,000 participants. In addition to the above aims, the NHS II provides the exposure data and initial case identification to examine many other important outcomes among young women.